An interview with Review Signal founder Kevin Ohashi

Kevin Ohashi founded Review Signal in 2011 as a source of honest web hosting reviews based on Twitter sentiment analysis, which was the subject of his master’s thesis at the University of Lund. Kevin also does comprehensive testing for an annual report on WordPress Hosting Performance Benchmarks, from entry level to enterprise accounts. To remain independent and unbiased, Review Signal’s revenue comes primarily from affiliate links. Kevin took some time out from his winter retreat in Thailand to chat with me about Review Signal’s past and future.

DK: You’ve been at this for seven years this month –– does it feel like a long time? Back in 2011 you said you hoped to become the Five Thirty Eight of the review industry. Are you still thinking about branching out into other consumer markets, or is consulting work keeping you busy?

KO: It feels like an eternity; it’s by far the longest project I’ve worked on. I’m still incredibly proud of it, but I think I fell short in my hope of transforming a dirty web hosting review industry (let alone going after reviews in a grander scheme). The shady pay-to-play affiliates still dominate the search engine rankings, and I’m rarely on the first page for big search terms like ‘web hosting reviews.’ It’s been beautiful to watch all the stories told in the data I’ve collected. I’ve been able to write some interesting pieces using it. I’ve also been able to use the platform to write about some of the shady stuff happening and try to call out the worst behavior in the industry. But I don’t think I’ve done enough.

As for branching out into different industries, it’s the most common question I get when I tell people what I do. Part of me thinks it would be a meaningful project to raise funding, get a team together and rebuild what I built to scale and go after a larger consumer market. The other part of me is quite burnt out after this long, and after going through the experience this long dreads that prospect. At this moment, I’m mostly head down doing consulting work that I barely get to even think about going after more ambitious plans.

DK: What else do you think you could do that would approach being “enough” to change the hosting industry the way you think it should change? What would that look like?

KO: My focus has always been on the consumer side, what’s good for them. I think Review Signal is a good start but it’s nowhere near perfect. The technology isn’t perfect, the method is limited in scope, and I can’t really commit full time to it anymore for financial reasons. I think having a bigger footprint on Google search and in the discussions that take place around the web would be the most impactful. I’m still certainly a tiny player by traffic volume for web hosting reviews and can’t really push as hard as the companies with bigger footprints.

If I was ranked higher, calling out bad behavior would be taken more seriously. I could take more of a watchdog role on how companies behave and potentially stop more bad behavior, or at least punish those who engage in deceptive tactics because of the visibility. A recent example is calling out Hostinger for their fake reviews which their CEO publicly said was OK and had no problem with. I got an email this week from one of their staff trying to partner with Review Signal. It’s clearly not impacted their behavior in any way.

If I had a bigger market share and mindshare, I would hope being told to stop cheating would actually cause companies to stop trying to cheat and reward companies that do well and are focused on improving their product and service for their customers.

DK: Elsewhere you’ve mentioned being pleased with the contraction of lower and mid-market hosting as premium managed platform-based hosting grows. Doesn’t that change indicate that accurate signals about host quality have had a transformative impact and will remain important?

KO: I’m not sure that the correlation there has any causation behind it. I think shared hosting is going to turn into a relic of an age where ‘webmasters’ were doing all sorts of things and installing a wide variety of software, snippets and code. It was a simple playground where one size fits all mentality was good enough. One plausible explanation is the one you posit – better specialized companies are displacing traditional lower end hosting by providing better service. Another plausible explanation is the web has opened up to a society that doesn’t want to tinker/code.

It’s easier to install WordPress and never write a line of code as you install plugin after plugin. Or use Wix/SquareSpace/etc to put together a beautiful looking site with all the tools at your fingertips. Companies are focusing on what users want, which is good. Whether those companies are actually good hosting companies — I’m not sure. There does seem to be a bit of a correlation with specialized hosting companies ranking better on average vs shared companies.

But the shared space has been eaten by a couple major companies that don’t have good reputations where a race to the bottom seems the most common outcome. I personally think specialization leads to better service on average for hosting companies, but that’s simply my opinion. The problem with success is many companies want to continue to grow and expand, and specialization may deteriorate over time. It will be interesting to see how the next 5-10 years play out with all the specialized hosting companies.

DK: Does your consulting focus on performance optimization, hosting, and sentiment analysis? The things you’ve learned about what works (and why) seem like they’d be of high value to some important players.

KO: I like to describe my consulting as helping businesses solve problems. I’ve definitely done performance optimization and hosting related consulting. I haven’t done much in the sentiment analysis space, but I’ve worked a lot with data in various capacities. The most interesting was working for an NHL team helping with hockey analytics. I also have done writing, marketing, technical evaluations, building all sorts of projects/ideas and more. I enjoy working with clients who need help solving business problems and appreciate a technical background and expertise.

I’ve worked very little in the web hosting space though. I’ve taken a few clients, but all of them have signed NDAs. I want to work with companies truly interested in improving their products and not simply wanting to try to attach my name or the Review Signal brand to themselves.

I am extremely selective when it comes to consulting in the web hosting space because of the issue of bias and perception. As far as taking what I’ve learned and bringing it to hosting companies, it’s actually the other way around. I’ve become friends with a lot of the best technical people in the WordPress hosting space, and they are teaching me constantly about new technology and best practices.

DK: Is there anyone doing similar work with sentiment analysis, or anything else that you’d consider complementary to what you’re doing?

KO: There used to be one site that did nearly exactly what I did, but manually. It was bought out by yet-another-garbage-review-site. It was incredibly complementary back when it ran because they didn’t have affiliate links and my data and theirs were incredibly close. I wrote about it, and it validated a lot of the work I did. Also Steven [Gliebe], who created the site, has become a friend and is a really great guy to talk to and commiserate together about the industry.

DK: Ten years ago it seemed like there were some really cool things being done with localizable social data, like patterns of behaviour, the spread of illnesses, and so on. We seem to have gone from the promise of open data to high levels of distrust and disinformation, but in the hosting space that’s always been the case. Would you say “open” doesn’t matter much if an honest signal is crowded out by an avalanche of noise?

KO: Data has gone into the walled gardens. I think we’ve gone the dystopian route towards everyone selling each other’s data off and treating people as the product. Twitter is the last semi-open social network that you can reliably pull data from. Facebook, Google, Apple, everyone is collecting tons of data and using it to sell better ads.

I can’t even get real search volumes anymore for keywords. Searching facebook is a nightmare to find anything meaningful that hasn’t been a notification in the last day or two. Open data is dying and we’re actively making it die quicker because it’s simply easier to consume in a pretty format than learning how to have control over it. Toss in maybe the most cynical era in history where “fake news” is more than a meme but a true belief, and we’re left in a state where even if you had data, people don’t want to believe it.

I actually really liked open data initiatives though coming from the government (not sure how much this is being pushed during this administration though). I was perhaps the first person to actually work with the CFPB’s data, and so there is a smaller glimmer of hope that at least in one critical area open data might live on and be impactful.

If we narrow it down to the hosting space specifically, open data has never really been the standard for the industry. Open data doesn’t matter if nobody sees, uses or respects it. I could have the best review site in the industry (which I believe I do have), but since barely anyone is aware it exists, it doesn’t really matter that much in the grand scheme of things.

DK: I’m not going to let you end on that note, even though I applaud and share your sense of pessimism/realism. At the end of the day, what keeps you going, and why is open data worth pursuing even if (or because?) the situation can seem hopeless? And what are you reading / watching for recreation and inspiration? Do books or television fit in with the nomad life?

KO: I’m just not ready to throw in the towel yet. While I may sound cynical or pessimistic, I actually feel like a bit of an optimist. I think the differences are made at the margins most of the time. Every interaction with open data in one way or another that is positive has a positive effect. In my case, providing accurate reviews about web hosting using open data to consumers is a small positive effect. It’s been a while since I’ve lived in DC, but I remember a lot of really passionate people trying to use government data to make things better.

Government data is especially important because it’s the basis for a lot of policy decisions, and the better we understand it, the better we can deploy resources to maximize impact. There’s a reason to be optimistic that open data will be used to help because serious people are funding it. One example that comes to mind is USAFacts which Steve Ballmer funded.

Right now I’m not reading anything sadly, I’ve been just working. But if you have any good non fiction recommendations, I’m all ears! As far as watching, I just finished BBC’s Sherlock, which I really enjoyed. I would say television/movies fit more into it than books, I admit, full of shame.